CN110399872B - Image processing method and device - Google Patents
Image processing method and device Download PDFInfo
- Publication number
- CN110399872B CN110399872B CN201910539227.6A CN201910539227A CN110399872B CN 110399872 B CN110399872 B CN 110399872B CN 201910539227 A CN201910539227 A CN 201910539227A CN 110399872 B CN110399872 B CN 110399872B
- Authority
- CN
- China
- Prior art keywords
- image
- rotation
- processed
- processing
- candidate
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 30
- 238000012545 processing Methods 0.000 claims abstract description 150
- 238000012937 correction Methods 0.000 claims description 47
- 238000012549 training Methods 0.000 claims description 24
- 238000004364 calculation method Methods 0.000 claims description 20
- 238000012015 optical character recognition Methods 0.000 claims description 16
- 238000003062 neural network model Methods 0.000 claims description 10
- 238000005516 engineering process Methods 0.000 claims description 9
- 230000015654 memory Effects 0.000 claims description 9
- 238000000034 method Methods 0.000 description 21
- 238000010586 diagram Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000004590 computer program Methods 0.000 description 2
- 238000003708 edge detection Methods 0.000 description 2
- 230000001413 cellular effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 239000004984 smart glass Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/60—Rotation of whole images or parts thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/24—Aligning, centring, orientation detection or correction of the image
- G06V10/242—Aligning, centring, orientation detection or correction of the image by image rotation, e.g. by 90 degrees
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Image Processing (AREA)
- Editing Of Facsimile Originals (AREA)
- Image Analysis (AREA)
Abstract
The application provides an image processing method and device, wherein the image processing method comprises the following steps: performing rotation processing on the image to be processed to obtain a rotation image; identifying text contents contained in candidate images formed by the image to be processed and the rotating image, and obtaining text contents corresponding to each candidate image; calculating a predicted probability value of the text content by using a statistical language model; and selecting a candidate image corresponding to the text content with the highest prediction probability value as a target image. According to the image processing method, the image to be processed is rotated to obtain the rotation image, and the text content contained in the candidate image formed by the image to be processed and the rotation image is identified and reasonably predicted at the fine granularity level, so that the target image is selected for archiving, the quality of image archiving is improved, and meanwhile, the image processing method has strong applicability.
Description
Technical Field
The present disclosure relates to the field of image processing technologies, and in particular, to an image processing method. The present application also relates to an image processing apparatus, a computing device, and a computer-readable storage medium.
Background
In many fields, electronic archiving is needed to facilitate unified management, query and analysis of paper documents, but when scanning paper documents, there may be different degrees of deflection problems of scanned document pictures due to equipment (scanners, etc.) or manual operation, such as deflecting scanned documents to the left or right by 90 or 180 degrees during scanning, and these deflected document pictures seriously affect archiving quality after electronic archiving and may affect subsequent other applications (e.g., contract comparison, contract structuring, etc.).
At present, a correction method for deflection pictures takes pixel points as analysis units, the picture is corrected at the pixel level, in the specific processing process, firstly, based on the assumption that text contents in document pictures are all aligned left, then, the distribution of text pixel points on a transverse axis in the pictures is counted, so that whether deflection conditions exist is judged, but the method has large limitation, and only pictures which strictly meet the assumption that the text contents in the pictures are aligned left can be processed, so that certain defects exist.
Disclosure of Invention
In view of the foregoing, embodiments of the present application provide an image processing method to solve the technical drawbacks in the prior art. Embodiments of the present application also provide an image processing apparatus, a computing device, and a computer-readable storage medium.
The application provides an image processing method, which comprises the following steps:
performing rotation processing on the image to be processed to obtain a rotation image;
identifying text contents contained in candidate images formed by the image to be processed and the rotating image, and obtaining text contents corresponding to each candidate image;
calculating a predicted probability value of the text content by using a statistical language model;
and selecting a candidate image corresponding to the text content with the highest prediction probability value as a target image.
Optionally, the rotating the image to be processed to obtain a rotated image includes:
performing rotation processing on the image to be processed according to a plurality of preset rotation angles to obtain a rotation image corresponding to each rotation angle;
wherein the rotation angle includes at least one of:
rotated 90 degrees clockwise, rotated 180 degrees clockwise, and rotated 270 degrees clockwise.
Optionally, the identifying text content included in the candidate image formed by the image to be processed and the rotation image, and obtaining text content corresponding to each candidate image includes:
and recognizing text contents contained in the candidate images by adopting an optical character recognition technology, and taking the text contents as the text contents corresponding to the candidate images.
Optionally, the statistical language model is trained in the following manner:
training a pre-constructed neural network model by using training corpus, and obtaining the statistical language model after training is completed;
the input of the statistical language model comprises text contents corresponding to the candidate images respectively, and the predicted probability value comprising the text contents is output.
Optionally, after the step of selecting the candidate image corresponding to the text content with the highest predicted probability value as the target image is performed, the method includes:
carrying out gray processing on the target image to obtain gray characteristics of the target image;
determining a feature edge region of the target image based on the gray scale features;
calculating a deflection angle of the target image according to the horizontal reference direction and the characteristic edge region;
judging whether the deflection angle is larger than a preset deflection angle threshold value or not;
and if so, carrying out deflection correction processing on the target image according to the correction angle corresponding to the deflection angle.
Optionally, the rotating the image to be processed to obtain a rotated image includes:
carrying out gray scale processing on the image to be processed to obtain gray scale characteristics of the image to be processed;
Determining a characteristic edge area of the image to be processed based on the gray scale characteristics;
calculating the deflection angle of the image to be processed according to the horizontal reference direction and the characteristic edge area;
calculating a plurality of rotation angles for performing rotation processing on the image to be processed based on the deflection angle;
and carrying out rotation processing on the image to be processed according to each rotation angle obtained through calculation, and obtaining a rotation image corresponding to each rotation angle.
Optionally, the step of performing rotation processing on the image to be processed, before performing the step of obtaining the rotation image, includes:
carrying out gray scale processing on the image to be processed to obtain a gray scale region of the image to be processed;
calculating a deflection angle of the image to be processed based on a horizontal reference direction and the gray scale region;
judging whether the deflection angle is larger than a preset deflection angle threshold value or not;
and if so, carrying out deflection correction processing on the image to be processed according to the correction angle corresponding to the deflection angle.
The present application also provides an image processing apparatus including:
the rotation processing module is configured to perform rotation processing on the image to be processed to obtain a rotation image;
the text content identification module is configured to identify text contents contained in candidate images formed by the image to be processed and the rotating image, and obtain text contents corresponding to the candidate images;
A text content prediction module configured to calculate a predicted probability value for the text content using a statistical language model;
and the target image selection module is configured to select a candidate image corresponding to the text content with the highest prediction probability value as the target image.
Optionally, the rotation processing module is specifically configured to perform rotation processing on the image to be processed according to a plurality of preset rotation angles, so as to obtain a rotation image corresponding to each rotation angle;
wherein the rotation angle includes at least one of: rotated 90 degrees clockwise, rotated 180 degrees clockwise, and rotated 270 degrees clockwise.
Optionally, the image processing apparatus further includes:
the gray processing module is configured to perform gray processing on the target image to obtain gray characteristics of the target image;
an edge region determination module configured to determine a feature edge region of the target image based on the gray scale features;
a deflection angle calculation module configured to calculate a deflection angle of the target image from a horizontal reference direction and the characteristic edge region;
a deflection angle judging module configured to judge whether the deflection angle is greater than a preset deflection angle threshold;
If yes, operating a correction processing module; the correction processing module is configured to perform deflection correction processing on the target image according to the correction angle corresponding to the deflection angle.
Optionally, the rotation processing module includes:
the gray processing submodule is configured to perform gray processing on the image to be processed to obtain gray characteristics of the image to be processed;
an edge region determination sub-module configured to determine a feature edge region of the image to be processed based on the gray scale features;
a deflection angle calculation sub-module configured to calculate a deflection angle of the image to be processed from a horizontal reference direction and the characteristic edge region;
a rotation angle calculation sub-module configured to calculate a plurality of rotation angles for performing rotation processing on the image to be processed based on the deflection angle;
and the rotation processing sub-module is configured to perform rotation processing on the image to be processed according to the calculated rotation angles to obtain a rotation image corresponding to each rotation angle.
Optionally, the image processing apparatus further includes:
the processing module is configured to perform gray processing on the image to be processed to obtain a gray area of the image to be processed;
A calculation module configured to calculate a deflection angle of the image to be processed based on a horizontal reference direction and the gray area;
the judging module is configured to judge whether the deflection angle is larger than a preset deflection angle threshold value or not;
if yes, operating a processing module; the processing module is configured to perform deflection correction processing on the image to be processed according to the correction angle corresponding to the deflection angle.
The present application also provides a computing device comprising:
a memory and a processor;
the memory is for storing computer-executable instructions, and the processor is for executing the computer-executable instructions:
performing rotation processing on the image to be processed to obtain a rotation image;
identifying text contents contained in candidate images formed by the image to be processed and the rotating image, and obtaining text contents corresponding to each candidate image;
calculating a predicted probability value of the text content by using a statistical language model;
and selecting a candidate image corresponding to the text content with the highest prediction probability value as a target image.
The present application also provides a computer-readable storage medium storing computer instructions that, when executed by a processor, implement the steps of the image processing method.
Compared with the prior art, the application has the following advantages:
the application provides an image processing method, which comprises the following steps: performing rotation processing on the image to be processed to obtain a rotation image; identifying text contents contained in candidate images formed by the image to be processed and the rotating image, and obtaining text contents corresponding to each candidate image; calculating a predicted probability value of the text content by using a statistical language model; and selecting a candidate image corresponding to the text content with the highest prediction probability value as a target image.
According to the image processing method, the image to be processed is rotated to obtain the rotated images covering different deflection angles, so that candidate images containing various deflection angles are formed together with the image to be processed, the text content contained in the candidate images is further identified and reasonably predicted on the fine granularity layer, the candidate images are selected to serve as target images to be archived on the basis of text content identification and prediction, the quality of image archiving is improved, and meanwhile, the image processing method has strong applicability.
Drawings
Fig. 1 is a process flow chart of an image processing method according to an embodiment of the present application;
FIG. 2 is a schematic illustration of a candidate image provided in an embodiment of the present application;
FIG. 3 is a schematic illustration of a document image provided in an embodiment of the present application;
fig. 4 is a schematic diagram of an image processing apparatus according to an embodiment of the present application;
fig. 5 is a block diagram of a computing device provided in an embodiment of the present application.
Detailed Description
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application. This application is, however, susceptible of embodiment in many other ways than those herein described and similar generalizations can be made by those skilled in the art without departing from the spirit of the application and the application is therefore not limited to the specific embodiments disclosed below.
The terminology used in the one or more embodiments of the specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the one or more embodiments of the specification. As used in this specification, one or more embodiments and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used in one or more embodiments of the present specification refers to and encompasses any or all possible combinations of one or more of the associated listed items.
It should be understood that, although the terms first, second, etc. may be used in one or more embodiments of this specification to describe various information, these information should not be limited by these terms. These terms are only used to distinguish one type of information from another. For example, a first may also be referred to as a second, and similarly, a second may also be referred to as a first, without departing from the scope of one or more embodiments of the present description. The word "if" as used herein may be interpreted as "at … …" or "at … …" or "responsive to a determination", depending on the context.
The application provides an image processing method, an image processing apparatus, a computing device, and a computer-readable storage medium. The following detailed description, together with the drawings of the embodiments provided herein, respectively, describes the steps of the method one by one.
An embodiment of an image processing method provided in the present application is as follows:
referring to fig. 1, a process flow chart of an image processing method provided in this embodiment is shown, referring to fig. 2, a schematic diagram of a candidate image provided in this embodiment is shown, and referring to fig. 3, a schematic diagram of a document image provided in this embodiment is shown.
Step S102, rotating the image to be processed to obtain a rotated image.
In the process of electronically archiving paper documents, the scanning archiving is generally performed manually, so that improper position emission of the paper documents often occurs in the scanning archiving process, so that document pictures after scanning archiving are deflected to different degrees, for example, the paper documents which are required to be placed in the forward direction are transversely scanned, so that the document images obtained after scanning are deflected by 90 degrees or 270 degrees from the images which are required to be archived, or the paper documents which are required to be placed in the forward direction are scanned after being inverted, so that the document images obtained after scanning are deflected by 180 degrees from the document images which are required to be archived, so that the archiving quality is affected, and the application of the document images after archiving is also affected.
In the image processing method provided by the embodiment of the application, in the process of archiving the document image, firstly, the image to be processed is rotated to obtain the rotated images covering different deflection angles, so that candidate images containing different deflection angles are formed together with the image to be processed, further, from the candidate images, the candidate images with the highest rationality of the contained text content are archived as the target image by identifying the text content contained in the candidate images and predicting the rationality of the text content, and the quality of image archiving is improved.
In a specific implementation, in an image archiving process of a paper document, rotation processing is performed on the image to be processed, which is needed to be archived, so as to obtain an image capable of covering multiple directions, and in an alternative implementation provided in the embodiment of the present application, the rotation processing is performed on the image to be processed by adopting the following manner:
performing rotation processing on the image to be processed according to a plurality of preset rotation angles to obtain a rotation image corresponding to each rotation angle;
wherein the rotation angle includes at least one of:
rotated 90 degrees clockwise, rotated 180 degrees clockwise, and rotated 270 degrees clockwise.
For example, as shown in fig. 2 (a), for a document image a of a paper document to be subjected to archiving processing, the rotation angles are set in advance at 3: rotating the document image A by 90 degrees clockwise, 180 degrees clockwise and 270 degrees clockwise, and performing rotation processing on the document image A, wherein the document image after the document image A is rotated by 90 degrees clockwise is used as a first rotation image, as shown in fig. 2 (b); taking the document image after the document image a is rotated 180 degrees clockwise as a second rotated image, as shown in fig. 2 (c); taking the document image after the document image a is rotated clockwise by 270 degrees as a third rotated image, as shown in fig. 2 (d);
The document image a shown in fig. 2 (a) can be regarded as a rotation image after the document image a rotates clockwise by 0 degrees, and it can be seen that the text directions of text contents contained in 4 rotation images obtained after the image rotation processing are respectively corresponding to different directions, and the document images in the 4 directions together form a candidate image (candidate image set); in practice, when the paper document is scanned in the archiving process, no matter the paper document is transversely placed or inverted, the document image with the final text direction being forward exists in candidate images formed by the 4 document images.
As described above, in practical application, besides that the paper document which should be placed in the forward direction is scanned transversely, resulting in a deflection of 90 degrees or 270 degrees between the document image obtained after scanning and the image which is expected to be archived, or the paper document which should be placed in the forward direction is scanned upside down, resulting in a deflection of 180 degrees between the document image obtained after scanning and the document image which is expected to be archived, in addition, there may be a situation that the paper document is placed obliquely during the paper document scanning process, in order to avoid that the scanning after the paper document is placed obliquely affects the document image archiving process, and in order to improve the quality of the image archiving, in an alternative embodiment provided in the embodiment of the present application, the rotation process is performed on the image to be processed in the following manner:
1) Carrying out gray scale processing on the image to be processed to obtain gray scale characteristics of the image to be processed;
2) Determining a characteristic edge area of the image to be processed based on the gray scale characteristics;
3) Calculating the deflection angle of the image to be processed according to the horizontal reference direction and the characteristic edge area;
4) Calculating a plurality of rotation angles for performing rotation processing on the image to be processed based on the deflection angle;
5) And carrying out rotation processing on the image to be processed according to each rotation angle obtained through calculation, and obtaining a rotation image corresponding to each rotation angle.
For example, a paper document is placed obliquely during scanning, and a document image a obtained after scanning is shown in fig. 3;
firstly, carrying out gray scale processing, namely binarization processing, on a document image A, and obtaining gray scale values of each pixel point in the document image A after gray scale processing;
secondly, detecting edge characteristics of the document image A in each direction by using an image edge detection algorithm (such as a Canny edge detection operator), and forming an integral edge area of the document image A;
thirdly, calculating the deflection angle (inclination angle) of the document image A according to the edge area of the whole document image A and the horizontal reference direction, wherein the calculated deflection angle of the document image A is 10 degrees clockwise deflection, as shown in fig. 3;
Then, from the inclination angle by which the document image a obtained by calculation is deflected clockwise by 10 degrees, the rotation angle of the document image a is determined: if the document image a is rotated back to the horizontal reference direction by rotation from 10 degrees clockwise, it is rotated 10 degrees counterclockwise, that is: the document image is rotated clockwise by 350 degrees, and the document image after the document image A is rotated according to the rotation angle is shown in fig. 2 (a);
there are 3 rotation angles of the preset image: rotated 90 degrees clockwise, rotated 180 degrees clockwise, rotated 270 degrees clockwise;
since the document image a itself has a deflection of 10 degrees clockwise, if the document image a needs to be rotated to a rotation angle of 90 degrees clockwise, it is only required to rotate by 80 degrees clockwise on the basis of the deflection of 10 degrees clockwise, and the rotated document image is as shown in fig. 2 (b);
by analogy, if the document image A needs to be rotated to a rotation angle of 180 degrees clockwise, the document image A only needs to be rotated by 170 degrees clockwise on the basis of existing clockwise 10-degree deflection, and the rotated document image is shown in the figure 2 (c);
if the document image A is required to be rotated to a rotation angle of 270 degrees clockwise, the document image A is required to be rotated to 260 degrees clockwise on the basis of the existing deflection of 10 degrees clockwise, and the rotated document image is shown in the figure 2 (d);
Finally, according to the determined 4 rotation angles of clockwise rotation by 80 degrees, clockwise rotation by 170 degrees, clockwise rotation by 260 degrees and clockwise rotation by 350 degrees, carrying out rotation processing on the document image A to obtain a first rotation image shown in fig. 2 (a), a second rotation image shown in fig. 2 (b), a third rotation image shown in fig. 2 (c) and a fourth rotation image shown in fig. 2 (d), wherein the text directions of text contents contained in the 4 rotation images respectively correspond to different directions, and the document images in the 4 directions jointly form a candidate image (candidate image set); in practice, when the paper document is scanned in the archiving process, under the condition that the paper document is placed obliquely, all the document images with the final text direction of forward directions exist in candidate images formed by the 4 document images.
In the above provided alternative embodiment, in the process of performing rotation processing on the image to be processed, the calculation of the rotation angle is combined to perform tilt deflection correction on the image to be processed, and in addition, before performing rotation processing on the image to be processed, tilt deflection correction may be performed on the image to be processed, so as to improve quality of image archiving.
1) Carrying out gray scale processing on the image to be processed to obtain a gray scale region of the image to be processed;
2) Calculating a deflection angle of the image to be processed based on a horizontal reference direction and the gray scale region;
3) Judging whether the deflection angle is larger than a preset deflection angle threshold value or not;
if yes, the deflection angle of the image to be processed is larger than the preset deflection angle threshold value, and the following step 4) is executed;
if not, indicating that the deflection angle of the image to be processed is smaller than or equal to the preset deflection angle threshold value, and not reaching the degree of deflection correction, and not processing;
4) And carrying out deflection correction processing on the image to be processed according to the correction angle corresponding to the deflection angle.
For example, a paper document is placed obliquely during scanning, and a document image a obtained after scanning is shown in fig. 3;
firstly, carrying out gray level processing, namely binarization processing, on a document image A, obtaining a gray level value of each pixel point in the document image A after gray level processing, and determining a gray level region of the document image A according to the gray level value of the pixel point;
secondly, calculating a deflection angle (inclination angle) of the document image A by using the gray scale area and the horizontal reference direction, wherein the deflection angle of the document image A obtained by calculation is 10 degrees of clockwise deflection;
Then judging whether the deflection angle of the document image A is larger than a preset deflection angle threshold (5 degrees of clockwise deflection or 5 degrees of anticlockwise deflection), and if so, correcting the deflection of the document image A by tilting the document image A, and rotating the document image A anticlockwise by 10 degrees (or 350 degrees of clockwise rotation) to restore the forward direction, wherein the correction angle is 10 degrees of anticlockwise rotation (or 350 degrees of clockwise rotation), and rotating the document image A according to the correction angle to obtain the document image shown in fig. 2 (a).
Step S104, identifying text contents contained in candidate images formed by the image to be processed and the rotation image, and obtaining text contents corresponding to each candidate image.
And after the image to be processed is subjected to rotation processing to obtain the rotation image, the candidate image is obtained, and then the text content contained in the candidate image is identified to obtain the text content corresponding to each candidate image. In an optional implementation manner provided in this embodiment of the present application, an optical character recognition (Optical Character Recognition, OCR) technology is specifically adopted to identify text content included in the candidate images, and the text content is used as text content corresponding to each of the candidate images.
For example, according to the candidate images composed of the 4-direction document images shown in fig. 2 obtained after the above rotation processing, the text content included in each candidate image is respectively identified by calling the packaged OCR service (the service of performing the text recognition by the OCR technology is packaged in advance), and the text content corresponding to each of the 4 candidate images is obtained after the identification is completed, specifically as follows:
1) In the first candidate image shown in fig. 2 (a), the first character on the left side is identified by OCR, and the identification result is a character similar to or close to the transverse (left-side transverse) character, or the character of the transverse "share" cannot be identified, and the identification result is a messy code;
similarly, after the three characters of the upper right, the lower left and the lower right in the first candidate image shown in fig. 2 (a) are identified by OCR, the identification result is similar or close to the characters of the transverse (left transverse) information, the body information and the information, or the three characters of the transverse information, the body information and the information cannot be identified, and the identification result is a messy code;
2) In the second candidate image shown in fig. 2 (b), recognition results of the 4 characters of the upper left, the upper right, the lower left and the lower right are sequentially: "body", "part", "message", "information";
3) In the third candidate image shown in fig. 2 (c), the first character on the left side is identified by OCR, and the identification result is a character similar to or close to the horizontal letter (right horizontal), or the character of the horizontal letter cannot be identified, and the identification result is a messy code;
similarly, after the three characters of the upper right, the lower left and the lower right in the third candidate image shown in fig. 2 (c) are identified by OCR, the identification result is a character similar to or close to the transverse (right transverse) body, the rest and the part, or the three characters of the transverse body, the rest and the part cannot be identified, and the identification result is a messy code;
4) In the fourth candidate image shown in fig. 2 (d), the first character on the left side is recognized by OCR, and the recognition result is a character similar to or close to the inverted "rest", or a character which is not recognized as the inverted "rest", and the recognition result is a messy code;
similarly, after the three characters of the upper right, the lower left and the lower right in the fourth candidate image shown in fig. 2 (d) are recognized by OCR, the recognition result is similar or close to the inverted "letter", "part", "body", or the three characters of the inverted "letter", "part", "body" cannot be recognized, and the recognition result is a messy code.
And step S106, calculating a predicted probability value of the text content by using a statistical language model.
After the text content corresponding to each candidate image is identified and obtained, the text content contained in the candidate image is identified by the same identification technology or identification algorithm according to each direction after the candidate image is rotated, so that the text direction of the text content obtained by identification corresponds to each direction, and correspondingly, part of the text content obtained by identification may exist as invalid (such as messy codes) or text content which has no meaning on a semantic level, wherein the statistical language model is used for predicting the prediction probability value of the text content, and the purpose of the statistical language model is to predict the rationality of the text content corresponding to each candidate image, wherein the rationality comprises the rationality of characters and the rationality of the characters.
Optionally, the statistical language model is trained in the following manner:
training a pre-constructed neural network model by using training corpus, and obtaining the statistical language model after training is completed;
in specific implementation, the neural network model may be constructed by using LSTM (Long Short-Term Memory) or other variants of LSTM, which is not limited thereto; the training corpus is a training sample of the neural network model obtained by training construction, and can be encyclopedic corpus, news corpus, text corpus in a specific field and the like;
The input of the statistical language model comprises text contents corresponding to the candidate images respectively, the output of the statistical language model comprises a predicted probability value of the text contents, and the predicted probability value output by the statistical language model is used for representing the rationality of the text contents of the candidate images.
Step S108, selecting a candidate image corresponding to the text content with the highest prediction probability value as a target image.
For the steps, calculating a predicted probability value for representing the rationality of the text content of the candidate image through the statistical language model, if the predicted probability value of the text content of the candidate image is higher, the rationality of the text content of the candidate image is higher, and correspondingly, the more characters are identified in the text content of the candidate image and the more accurate the identification is, so that the possibility that the text content of the candidate image is forward is also higher; conversely, if the predicted probability value of the text content of the candidate image is lower, the rationality of the text content of the candidate image is lower, and accordingly, the fewer characters are recognized in the text content of the candidate image and the more inaccurate recognition is, so that the probability that the text content of the candidate image is forward is also lower. In this embodiment of the present application, a candidate image corresponding to the text content with the highest prediction probability value is used as the target image, where the target image is the target image with the text content being the forward direction in the candidate image.
For example, as shown in 4 candidate images in fig. 2, recognition results of 4 characters of "part", "rest", "body" and "letter" in the lateral direction (left lateral direction) included in the first candidate image in fig. 2 (a) are input into the LSTM model to make predictions, and a prediction probability value for rationality of the recognition results is 15%;
the recognition results of the 4 characters, namely, the positive body, the positive information and the positive information, contained in the second candidate image shown in the figure 2 (b) are input into an LSTM model for prediction, and the prediction probability value of the rationality of the recognition result is 98%;
the recognition results of 4 characters of "letter", "body", "message" and "copy" in the transverse direction (right-side transverse direction) contained in the third candidate image shown in fig. 2 (c) are input into the LSTM model for prediction, and the prediction probability value of the rationality of the recognition result is output to be 17%;
the recognition results of the 4 characters of inverted "information", "letter", "identity" and "body" contained in the fourth candidate image shown in fig. 2 (d) are input into the LSTM model for prediction, and the prediction probability value of the rationality of the recognition result is output to be 9%;
as can be seen from the predicted probability value output by the LSTM model, the probability that the character content included in the second candidate image shown in fig. 2 (b) is forward is greatest, and therefore, the second candidate image shown in fig. 2 (b) is taken as the corrected target image of the original document image a.
As described above, in the above-mentioned alternative embodiments provided in the embodiments of the present application, before performing rotation processing on the image to be processed, deflection correction is performed on the image to be processed, and in addition, after determining a target image in the candidate image, deflection correction may be performed on the target image, which may also improve quality of image archiving, and specifically implement the following steps:
1) Carrying out gray processing on the target image to obtain gray characteristics of the target image;
2) Determining a feature edge region of the target image based on the gray scale features;
3) Calculating a deflection angle of the target image according to the horizontal reference direction and the characteristic edge region;
4) Judging whether the deflection angle is larger than a preset deflection angle threshold value or not;
if yes, the deflection angle of the target image is larger than the preset deflection angle threshold, and the following step 5) is executed;
if not, indicating that the deflection angle of the target image is smaller than or equal to the preset deflection angle threshold, and not reaching the degree of deflection correction, and not processing;
5) And carrying out deflection correction processing on the target image according to the correction angle corresponding to the deflection angle.
According to the image processing method, the image to be processed is rotated to obtain the rotated images covering different deflection angles, so that candidate images containing various deflection angles are formed together with the image to be processed, the text content contained in the candidate images is further identified and reasonably predicted on the fine granularity layer, the candidate images are selected to serve as target images to be archived on the basis of text content identification and prediction, the quality of image archiving is improved, and meanwhile, the image processing method has strong applicability.
An embodiment of an image processing apparatus provided in the present application is as follows:
in the above-described embodiments, an image processing method is provided, and accordingly, an image processing apparatus is also provided, and is described below with reference to the accompanying drawings.
Referring to fig. 4, a schematic diagram of an embodiment of an image processing apparatus provided in the present application is shown.
Since the apparatus embodiments are substantially similar to the method embodiments, the description is relatively simple, and reference should be made to the corresponding descriptions of the method embodiments provided above for relevant parts. The device embodiments described below are merely illustrative.
The application provides an image processing apparatus including:
a rotation processing module 402 configured to perform rotation processing on an image to be processed to obtain a rotation image;
a text content recognition module 404, configured to recognize text content contained in the candidate images composed of the image to be processed and the rotation image, and obtain text content corresponding to each candidate image;
a text content prediction module 406 configured to calculate a predicted probability value for the text content using a statistical language model;
the target image selection module 408 is configured to select, as the target image, a candidate image corresponding to the text content having the highest prediction probability value.
Optionally, the rotation processing module 402 is specifically configured to perform rotation processing on the image to be processed according to a plurality of preset rotation angles, so as to obtain a rotation image corresponding to each rotation angle;
wherein the rotation angle includes at least one of: rotated 90 degrees clockwise, rotated 180 degrees clockwise, and rotated 270 degrees clockwise.
Optionally, the text content recognition module 404 is specifically configured to recognize text content included in the candidate images as text content corresponding to each of the candidate images by using an optical character recognition technology.
Optionally, the statistical language model is trained in the following manner: training a pre-constructed neural network model by using training corpus, and obtaining the statistical language model after training is completed;
the input of the statistical language model comprises text contents corresponding to the candidate images respectively, and the predicted probability value comprising the text contents is output.
Optionally, the image processing apparatus further includes:
the gray processing module is configured to perform gray processing on the target image to obtain gray characteristics of the target image;
an edge region determination module configured to determine a feature edge region of the target image based on the gray scale features;
a deflection angle calculation module configured to calculate a deflection angle of the target image from a horizontal reference direction and the characteristic edge region;
a deflection angle judging module configured to judge whether the deflection angle is greater than a preset deflection angle threshold;
if yes, operating a correction processing module; the correction processing module is configured to perform deflection correction processing on the target image according to the correction angle corresponding to the deflection angle.
Optionally, the rotation processing module 402 includes:
the gray processing submodule is configured to perform gray processing on the image to be processed to obtain gray characteristics of the image to be processed;
an edge region determination sub-module configured to determine a feature edge region of the image to be processed based on the gray scale features;
a deflection angle calculation sub-module configured to calculate a deflection angle of the image to be processed from a horizontal reference direction and the characteristic edge region;
a rotation angle calculation sub-module configured to calculate a plurality of rotation angles for performing rotation processing on the image to be processed based on the deflection angle;
and the rotation processing sub-module is configured to perform rotation processing on the image to be processed according to the calculated rotation angles to obtain a rotation image corresponding to each rotation angle.
Optionally, the image processing apparatus further includes:
the processing module is configured to perform gray processing on the image to be processed to obtain a gray area of the image to be processed;
a calculation module configured to calculate a deflection angle of the image to be processed based on a horizontal reference direction and the gray area;
The judging module is configured to judge whether the deflection angle is larger than a preset deflection angle threshold value or not;
if yes, operating a processing module; the processing module is configured to perform deflection correction processing on the image to be processed according to the correction angle corresponding to the deflection angle.
An embodiment of a computing device provided herein is as follows:
fig. 5 is a block diagram illustrating a configuration of a computing device 500 according to an embodiment of the present description. The components of the computing device 500 include, but are not limited to, a memory 510 and a processor 520. Processor 520 is coupled to memory 510 via bus 530 and database 550 is used to hold data.
Computing device 500 also includes access device 540, access device 540 enabling computing device 500 to communicate via one or more networks 560. Examples of such networks include the Public Switched Telephone Network (PSTN), a Local Area Network (LAN), a Wide Area Network (WAN), a Personal Area Network (PAN), or a combination of communication networks such as the internet. The access device 540 may include one or more of any type of network interface, wired or wireless (e.g., a Network Interface Card (NIC)), such as an IEEE802.11 Wireless Local Area Network (WLAN) wireless interface, a worldwide interoperability for microwave access (Wi-MAX) interface, an ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a bluetooth interface, a Near Field Communication (NFC) interface, and so forth.
In one embodiment of the present description, the above-described components of computing device 500, as well as other components not shown in FIG. 5, may also be connected to each other, such as by a bus. It should be understood that the block diagram of the computing device shown in FIG. 5 is for exemplary purposes only and is not intended to limit the scope of the present description. Those skilled in the art may add or replace other components as desired.
Computing device 500 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., tablet, personal digital assistant, laptop, notebook, netbook, etc.), mobile phone (e.g., smart phone), wearable computing device (e.g., smart watch, smart glasses, etc.), or other type of mobile device, or a stationary computing device such as a desktop computer or PC. Computing device 500 may also be a mobile or stationary server.
The present application provides a computing device comprising a memory 510, a processor 520, and computer instructions stored on the memory and executable on the processor, the processor 520 for executing computer executable instructions to:
Performing rotation processing on the image to be processed to obtain a rotation image;
identifying text contents contained in candidate images formed by the image to be processed and the rotating image, and obtaining text contents corresponding to each candidate image;
calculating a predicted probability value of the text content by using a statistical language model;
and selecting a candidate image corresponding to the text content with the highest prediction probability value as a target image.
Optionally, the rotating the image to be processed to obtain a rotated image includes:
performing rotation processing on the image to be processed according to a plurality of preset rotation angles to obtain a rotation image corresponding to each rotation angle;
wherein the rotation angle includes at least one of:
rotated 90 degrees clockwise, rotated 180 degrees clockwise, and rotated 270 degrees clockwise.
Optionally, the identifying text content included in the candidate image formed by the image to be processed and the rotation image, and obtaining text content corresponding to each candidate image includes:
and recognizing text contents contained in the candidate images by adopting an optical character recognition technology, and taking the text contents as the text contents corresponding to the candidate images.
Optionally, the statistical language model is trained in the following manner:
Training a pre-constructed neural network model by using training corpus, and obtaining the statistical language model after training is completed;
the input of the statistical language model comprises text contents corresponding to the candidate images respectively, and the predicted probability value comprising the text contents is output.
Optionally, after the candidate image corresponding to the text content with the highest predicted probability value is selected and executed as the target image instruction, the processor 520 is further configured to execute the following computer executable instructions:
carrying out gray processing on the target image to obtain gray characteristics of the target image;
determining a feature edge region of the target image based on the gray scale features;
calculating a deflection angle of the target image according to the horizontal reference direction and the characteristic edge region;
judging whether the deflection angle is larger than a preset deflection angle threshold value or not;
and if so, carrying out deflection correction processing on the target image according to the correction angle corresponding to the deflection angle.
Optionally, the rotating the image to be processed to obtain a rotated image includes:
carrying out gray scale processing on the image to be processed to obtain gray scale characteristics of the image to be processed;
Determining a characteristic edge area of the image to be processed based on the gray scale characteristics;
calculating the deflection angle of the image to be processed according to the horizontal reference direction and the characteristic edge area;
calculating a plurality of rotation angles for performing rotation processing on the image to be processed based on the deflection angle;
and carrying out rotation processing on the image to be processed according to each rotation angle obtained through calculation, and obtaining a rotation image corresponding to each rotation angle.
Optionally, before the rotation processing is performed on the image to be processed to obtain the rotation image instruction, the processor 520 is further configured to execute the following computer executable instructions:
carrying out gray scale processing on the image to be processed to obtain a gray scale region of the image to be processed;
calculating a deflection angle of the image to be processed based on a horizontal reference direction and the gray scale region;
judging whether the deflection angle is larger than a preset deflection angle threshold value or not;
and if so, carrying out deflection correction processing on the image to be processed according to the correction angle corresponding to the deflection angle.
An embodiment of a computer readable storage medium provided in the present application is as follows:
the present application provides a computer readable storage medium storing computer instructions that when executed by a processor are configured to:
Performing rotation processing on the image to be processed to obtain a rotation image;
identifying text contents contained in candidate images formed by the image to be processed and the rotating image, and obtaining text contents corresponding to each candidate image;
calculating a predicted probability value of the text content by using a statistical language model;
and selecting a candidate image corresponding to the text content with the highest prediction probability value as a target image.
Optionally, the rotating the image to be processed to obtain a rotated image includes:
performing rotation processing on the image to be processed according to a plurality of preset rotation angles to obtain a rotation image corresponding to each rotation angle;
wherein the rotation angle includes at least one of:
rotated 90 degrees clockwise, rotated 180 degrees clockwise, and rotated 270 degrees clockwise.
Optionally, the identifying text content included in the candidate image formed by the image to be processed and the rotation image, and obtaining text content corresponding to each candidate image includes:
and recognizing text contents contained in the candidate images by adopting an optical character recognition technology, and taking the text contents as the text contents corresponding to the candidate images.
Optionally, the statistical language model is trained in the following manner:
Training a pre-constructed neural network model by using training corpus, and obtaining the statistical language model after training is completed;
the input of the statistical language model comprises text contents corresponding to the candidate images respectively, and the predicted probability value comprising the text contents is output.
Optionally, after the step of selecting the candidate image corresponding to the text content with the highest predicted probability value as the target image is performed, the method includes:
carrying out gray processing on the target image to obtain gray characteristics of the target image;
determining a feature edge region of the target image based on the gray scale features;
calculating a deflection angle of the target image according to the horizontal reference direction and the characteristic edge region;
judging whether the deflection angle is larger than a preset deflection angle threshold value or not;
and if so, carrying out deflection correction processing on the target image according to the correction angle corresponding to the deflection angle.
Optionally, the rotating the image to be processed to obtain a rotated image includes:
carrying out gray scale processing on the image to be processed to obtain gray scale characteristics of the image to be processed;
determining a characteristic edge area of the image to be processed based on the gray scale characteristics;
Calculating the deflection angle of the image to be processed according to the horizontal reference direction and the characteristic edge area;
calculating a plurality of rotation angles for performing rotation processing on the image to be processed based on the deflection angle;
and carrying out rotation processing on the image to be processed according to each rotation angle obtained through calculation, and obtaining a rotation image corresponding to each rotation angle.
Optionally, the step of performing rotation processing on the image to be processed, before performing the step of obtaining the rotation image, includes:
carrying out gray scale processing on the image to be processed to obtain a gray scale region of the image to be processed;
calculating a deflection angle of the image to be processed based on a horizontal reference direction and the gray scale region;
judging whether the deflection angle is larger than a preset deflection angle threshold value or not;
and if so, carrying out deflection correction processing on the image to be processed according to the correction angle corresponding to the deflection angle.
The above is an exemplary version of a computer-readable storage medium of the present embodiment. It should be noted that, the technical solution of the storage medium and the technical solution of the image processing method belong to the same concept, and details of the technical solution of the storage medium which are not described in detail can be referred to the description of the technical solution of the image processing method.
The computer instructions include computer program code that may be in source code form, object code form, executable file or some intermediate form, etc. The computer readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a U disk, a removable hard disk, a magnetic disk, an optical disk, a computer Memory, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), an electrical carrier signal, a telecommunications signal, a software distribution medium, and so forth. It should be noted that the computer readable medium contains content that can be appropriately scaled according to the requirements of jurisdictions in which such content is subject to legislation and patent practice, such as in certain jurisdictions in which such content is subject to legislation and patent practice, the computer readable medium does not include electrical carrier signals and telecommunication signals.
It should be noted that, for the sake of simplicity of description, the foregoing method embodiments are all expressed as a series of combinations of actions, but it should be understood by those skilled in the art that the present application is not limited by the order of actions described, as some steps may be performed in other order or simultaneously in accordance with the present application. Further, those skilled in the art will also appreciate that the embodiments described in the specification are all preferred embodiments, and that the acts and modules referred to are not necessarily all necessary for the present application.
In the foregoing embodiments, the descriptions of the embodiments are emphasized, and for parts of one embodiment that are not described in detail, reference may be made to the related descriptions of other embodiments.
The above-disclosed preferred embodiments of the present application are provided only as an aid to the elucidation of the present application. Alternative embodiments are not intended to be exhaustive or to limit the invention to the precise form disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the application and the practical application, to thereby enable others skilled in the art to best understand and utilize the application. This application is to be limited only by the claims and the full scope and equivalents thereof.
Claims (13)
1. An image processing method, comprising:
performing rotation processing on the image to be processed according to a plurality of preset rotation angles to obtain a rotation image corresponding to each rotation angle;
identifying text contents contained in candidate images formed by the image to be processed and the rotating image, and obtaining text contents corresponding to each candidate image;
calculating a prediction probability value of the text content by using a statistical language model, wherein the statistical language model is obtained by training a pre-constructed neural network model by using a training corpus, the prediction probability value is used for representing the rationality of the text content of the candidate image, the input of the statistical language model comprises the text content corresponding to each candidate image, and the prediction probability value comprising the text content is output;
And selecting a candidate image corresponding to the text content with the highest prediction probability value as a target image.
2. The image processing method according to claim 1, wherein the rotation angle includes at least one of:
rotated 90 degrees clockwise, rotated 180 degrees clockwise, and rotated 270 degrees clockwise.
3. The image processing method according to claim 1, wherein the identifying text content included in the candidate images composed of the image to be processed and the rotated image, obtaining text content corresponding to each candidate image, includes:
and recognizing text contents contained in the candidate images by adopting an optical character recognition technology, and taking the text contents as the text contents corresponding to the candidate images.
4. The image processing method according to claim 1, wherein after the step of selecting, as the target image, the candidate image corresponding to the text content having the highest predicted probability value, comprises:
carrying out gray processing on the target image to obtain gray characteristics of the target image;
determining a feature edge region of the target image based on the gray scale features;
calculating a deflection angle of the target image according to the horizontal reference direction and the characteristic edge region;
Judging whether the deflection angle is larger than a preset deflection angle threshold value or not;
and if so, carrying out deflection correction processing on the target image according to the correction angle corresponding to the deflection angle.
5. The image processing method according to claim 1, wherein the rotating processing of the image to be processed according to a plurality of rotation angles set in advance to obtain a rotation image corresponding to each rotation angle includes:
carrying out gray scale processing on the image to be processed to obtain gray scale characteristics of the image to be processed;
determining a characteristic edge area of the image to be processed based on the gray scale characteristics;
calculating the deflection angle of the image to be processed according to the horizontal reference direction and the characteristic edge area;
calculating a plurality of rotation angles for performing rotation processing on the image to be processed based on the deflection angle;
and carrying out rotation processing on the image to be processed according to each rotation angle obtained through calculation, and obtaining a rotation image corresponding to each rotation angle.
6. The image processing method according to claim 1, wherein the step of performing rotation processing on the image to be processed according to a plurality of rotation angles set in advance to obtain rotation images corresponding to the respective rotation angles includes, before performing:
Carrying out gray scale processing on the image to be processed to obtain a gray scale region of the image to be processed;
calculating a deflection angle of the image to be processed based on a horizontal reference direction and the gray scale region;
judging whether the deflection angle is larger than a preset deflection angle threshold value or not;
and if so, carrying out deflection correction processing on the image to be processed according to the correction angle corresponding to the deflection angle.
7. An image processing apparatus, comprising:
the rotation processing module is configured to perform rotation processing on the image to be processed according to a plurality of preset rotation angles to obtain a rotation image corresponding to each rotation angle;
the text content identification module is configured to identify text contents contained in candidate images formed by the image to be processed and the rotating image, and obtain text contents corresponding to the candidate images;
a text content prediction module configured to calculate a prediction probability value of the text content by using a statistical language model, wherein the statistical language model is obtained by training a pre-constructed neural network model by using a training corpus, the prediction probability value is used for representing the rationality of the text content of the candidate image, the input of the statistical language model comprises the text content corresponding to each candidate image, and the output of the statistical language model comprises the prediction probability value of the text content;
And the target image selection module is configured to select a candidate image corresponding to the text content with the highest prediction probability value as the target image.
8. The image processing apparatus according to claim 7, wherein the rotation processing module is specifically configured such that the rotation angle comprises at least one of: rotated 90 degrees clockwise, rotated 180 degrees clockwise, and rotated 270 degrees clockwise.
9. The image processing apparatus according to claim 7, further comprising:
the gray processing module is configured to perform gray processing on the target image to obtain gray characteristics of the target image;
an edge region determination module configured to determine a feature edge region of the target image based on the gray scale features;
a deflection angle calculation module configured to calculate a deflection angle of the target image from a horizontal reference direction and the characteristic edge region;
a deflection angle judging module configured to judge whether the deflection angle is greater than a preset deflection angle threshold;
if yes, operating a correction processing module; the correction processing module is configured to perform deflection correction processing on the target image according to the correction angle corresponding to the deflection angle.
10. The image processing apparatus according to claim 7, wherein the rotation processing module includes:
the gray processing submodule is configured to perform gray processing on the image to be processed to obtain gray characteristics of the image to be processed;
an edge region determination sub-module configured to determine a feature edge region of the image to be processed based on the gray scale features;
a deflection angle calculation sub-module configured to calculate a deflection angle of the image to be processed from a horizontal reference direction and the characteristic edge region;
a rotation angle calculation sub-module configured to calculate a plurality of rotation angles for performing rotation processing on the image to be processed based on the deflection angle;
and the rotation processing sub-module is configured to perform rotation processing on the image to be processed according to the calculated rotation angles to obtain a rotation image corresponding to each rotation angle.
11. The image processing apparatus according to claim 7, further comprising:
the processing module is configured to perform gray processing on the image to be processed to obtain a gray area of the image to be processed;
a calculation module configured to calculate a deflection angle of the image to be processed based on a horizontal reference direction and the gray area;
The judging module is configured to judge whether the deflection angle is larger than a preset deflection angle threshold value or not;
if yes, operating a processing module; the processing module is configured to perform deflection correction processing on the image to be processed according to the correction angle corresponding to the deflection angle.
12. A computing device, comprising:
a memory and a processor;
the memory is for storing computer-executable instructions, and the processor is for executing the computer-executable instructions:
performing rotation processing on the image to be processed according to a plurality of preset rotation angles to obtain a rotation image corresponding to each rotation angle;
identifying text contents contained in candidate images formed by the image to be processed and the rotating image, and obtaining text contents corresponding to each candidate image;
calculating a prediction probability value of the text content by using a statistical language model, wherein the statistical language model is obtained by training a pre-constructed neural network model by using a training corpus, the prediction probability value is used for representing the rationality of the text content of the candidate image, the input of the statistical language model comprises the text content corresponding to each candidate image, and the prediction probability value comprising the text content is output;
And selecting a candidate image corresponding to the text content with the highest prediction probability value as a target image.
13. A computer-readable storage medium storing computer instructions which, when executed by a processor, implement the steps of the image processing method of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910539227.6A CN110399872B (en) | 2019-06-20 | 2019-06-20 | Image processing method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910539227.6A CN110399872B (en) | 2019-06-20 | 2019-06-20 | Image processing method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110399872A CN110399872A (en) | 2019-11-01 |
| CN110399872B true CN110399872B (en) | 2023-04-28 |
Family
ID=68324187
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910539227.6A Active CN110399872B (en) | 2019-06-20 | 2019-06-20 | Image processing method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110399872B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115941859A (en) * | 2021-08-16 | 2023-04-07 | 广州视源电子科技股份有限公司 | Image direction adjusting method and device, storage medium and electronic equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2011008549A (en) * | 2009-06-25 | 2011-01-13 | Sharp Corp | Image processor, image reader, multifunctional machine, image processing method, program, and recording medium |
| CN103714327A (en) * | 2013-12-30 | 2014-04-09 | 上海合合信息科技发展有限公司 | Method and system for correcting image direction |
| CN109376658A (en) * | 2018-10-26 | 2019-02-22 | 信雅达系统工程股份有限公司 | A kind of OCR method based on deep learning |
| CN109685052A (en) * | 2018-12-06 | 2019-04-26 | 泰康保险集团股份有限公司 | Method for processing text images, device, electronic equipment and computer-readable medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105550633B (en) * | 2015-10-30 | 2018-12-11 | 小米科技有限责任公司 | Area recognizing method and device |
-
2019
- 2019-06-20 CN CN201910539227.6A patent/CN110399872B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2011008549A (en) * | 2009-06-25 | 2011-01-13 | Sharp Corp | Image processor, image reader, multifunctional machine, image processing method, program, and recording medium |
| CN103714327A (en) * | 2013-12-30 | 2014-04-09 | 上海合合信息科技发展有限公司 | Method and system for correcting image direction |
| CN109376658A (en) * | 2018-10-26 | 2019-02-22 | 信雅达系统工程股份有限公司 | A kind of OCR method based on deep learning |
| CN109685052A (en) * | 2018-12-06 | 2019-04-26 | 泰康保险集团股份有限公司 | Method for processing text images, device, electronic equipment and computer-readable medium |
Non-Patent Citations (2)
| Title |
|---|
| 国土部门电力档案系统实现及研究;马青浙等;《网络财富》;20080701(第13期);全文 * |
| 图像处理技术在档案文本数字化中的应用;吴强等;《兰台世界》;20060416(第07期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110399872A (en) | 2019-11-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109344884B (en) | Media information classification method, method and device for training picture classification model | |
| CN109784181B (en) | Picture watermark identification method, device, equipment and computer readable storage medium | |
| RU2640296C1 (en) | Method and device for determining document suitability for optical character recognition (ocr) on server | |
| CN110222694B (en) | Image processing method, image processing device, electronic equipment and computer readable medium | |
| WO2022193523A1 (en) | Image processing method and apparatus, device, and storage medium | |
| CN111507957B (en) | Identity card picture conversion method and device, computer equipment and storage medium | |
| CN110781874A (en) | Method for converting electronic menu into menu of cooking machine and menu making terminal | |
| US9117132B2 (en) | System and method facilitating designing of classifier while recognizing characters in a video | |
| CN108197638B (en) | Method and device for classifying sample to be evaluated | |
| CN114494775A (en) | Video segmentation method, device, equipment and storage medium | |
| US8773733B2 (en) | Image capture device for extracting textual information | |
| CN110399872B (en) | Image processing method and device | |
| CN110210467B (en) | Formula positioning method of text image, image processing device and storage medium | |
| CN113792659B (en) | Document identification method and device and electronic equipment | |
| CN115131590A (en) | Training method of target detection model, target detection method and related equipment | |
| CN113657370B (en) | Character recognition method and related equipment thereof | |
| US8768058B2 (en) | System for extracting text from a plurality of captured images of a document | |
| US8908970B2 (en) | Textual information extraction method using multiple images | |
| CN111145202A (en) | Model generation method, image processing method, device, equipment and storage medium | |
| CN116665228B (en) | Image processing method and device | |
| CN114399497A (en) | Text image quality detection method and device, computer equipment and storage medium | |
| CN114663418A (en) | Image processing method and device, storage medium and electronic equipment | |
| CN114494686A (en) | Text image correction method, text image correction device, electronic equipment and storage medium | |
| CN110020728B (en) | Service model reinforcement learning method and device | |
| CN110991270B (en) | Text recognition method, device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20201010 Address after: Cayman Enterprise Centre, 27 Hospital Road, George Town, Grand Cayman Islands Applicant after: Advanced innovation technology Co.,Ltd. Address before: A four-storey 847 mailbox in Grand Cayman Capital Building, British Cayman Islands Applicant before: Alibaba Group Holding Ltd. Effective date of registration: 20201010 Address after: Cayman Enterprise Centre, 27 Hospital Road, George Town, Grand Cayman Islands Applicant after: Innovative advanced technology Co.,Ltd. Address before: Cayman Enterprise Centre, 27 Hospital Road, George Town, Grand Cayman Islands Applicant before: Advanced innovation technology Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |